Antifungal Activity of the Noncytotoxic Human Peptide Hepcidin 20against Fluconazole-Resistant Candida glabrata in Human VaginalFluid

Gaetano Del Gaudio,a Lisa Lombardi,a Giuseppantonio Maisetta,b Semih Esin,b Giovanna Batoni,b Maurizio Sanguinetti,c

Sonia Senesi,a Arianna Tavantia

Dipartimento di Biologia, Università di Pisa, Pisa, Italya; Dipartimento di Ricerca Traslazionale e delle Nuove Tecnologie in Medicina e Chirurgia, Università di Pisa, Pisa,Italyb; Istituto di Microbiologia, Università Cattolica del Sacro Cuore, Roma, Italyc

Vaginal infections caused by Candida glabrata are difficult to eradicate due to this species’ scarce susceptibility to azoles. Previ-ous studies have shown that the human cationic peptide hepcidin 20 (Hep-20) exerts fungicidal activity in sodium phosphatebuffer against a panel of C. glabrata clinical isolates with different levels of susceptibility to fluconazole. In addition, the activityof the peptide was potentiated under acidic conditions, suggesting an application in the topical treatment of vaginal infections.To investigate whether the peptide activity could be maintained in biological fluids, in this study the antifungal activity ofHep-20 was evaluated by a killing assay in (i) a vaginal fluid simulant (VFS) and in (ii) human vaginal fluid (HVF) collected fromthree healthy donors. The results obtained indicated that the activity of the peptide was maintained in VFS and HVF supple-mented with EDTA. Interestingly, the fungicidal activity of Hep-20 was enhanced in HVF compared to that observed in VFS,with a minimal fungicidal concentration of 25 �� for all donors. No cytotoxic effect on human cells was exerted by Hep-20 atconcentrations ranging from 6.25 to 100 ��, as shown by 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-car-boxanilide tetrazolium salt (XTT) reduction assay and propidium iodide staining. A piece of indirect evidence of Hep-20 stabil-ity was also obtained from coincubation experiments of the peptide with HVF at 37°C for 90 min and for 24 h. Collectively, theseresults indicate that this peptide should be further studied as a novel therapeutic agent for the topical treatment of vaginal C.glabrata infections.

In the last decade there have been increasing reports of vaginitisdue to non-albicans Candida species, which now accounts for

20% of Candida vaginal infections (1). Candida glabrata hasemerged as an important fungal pathogen, and its clinical rele-vance is partially due to its intrinsic or rapidly acquired resistanceto azole antifungal drugs, such as fluconazole, commonly used inthe treatment of mucosal infections (1, 2). It has been postulatedthat the rise in C. glabrata prevalence from recurrent vulvovaginalcandidosis observed in the last few decades is related to the wide-spread use of fluconazole, to which this species is less sensitivethan C. albicans and other non-albicans Candida species (1, 2).Recent reports have revealed an increase of C. glabrata clinicalisolates that display resistance not only to azoles but also to echi-nocandins (3, 4). This aspect complicates the management of C.glabrata vaginal infections, which, as a result, are difficult to erad-icate and often end up in recurrent episodes (1).

Considering the limitations of the currently available antifun-gals, it has become mandatory to identify new classes of antimi-crobial compounds to provide an effective antifungal strategy fora broad range of fungal pathogens, including C. glabrata.

In this regard, natural anti-infective agents, such as antimicro-bial peptides (AMPs), represent a promising approach due to theirwide-spectrum activity, rapid mode of action, and microbicidaleffect on multiresistant microorganisms (5). Interactions betweenAMPs and the microbial surface play a crucial role in determiningthe activity of membrane-active AMPs. The positive charge of thevast majority of AMPs allows the initial interaction with the neg-atively charged cell wall and, later, with negatively charged phos-pholipids in the inner membrane. Following interaction with the

lipid bilayer, most AMPs cause displacement of lipids, alterationof membrane structure, and membrane perforation (5).

Structurally different AMPs, obtained from different organ-isms, including amphibians, plants, and mammals, have beenproven to exert antifungal activity versus clinically relevant Can-dida species (6). However, C. glabrata has been shown to bescarcely susceptible to a wide spectrum of AMPs of different der-ivations; among these are histatin, cathelicidins, defensins, andmagainin (7–10).

We have previously shown that the human liver-derived pep-tide hepcidin 20 (Hep-20) exerts antifungal activity against C.glabrata clinical isolates characterized by different levels of suscep-tibility to fluconazole (11). In addition, the fungicidal effect ex-erted by Hep-20 on C. glabrata isolates in sodium phosphate buf-fer (SPB) was potentiated under acidic conditions (pH 5.0) (11),according to a proposed mechanism by which, at acid pH, proto-nated basic residues of Hep-20 facilitate the interaction with neg-atively charged fungal surfaces, as already reported for other pep-

Received 29 April 2013 Returned for modification 3 June 2013Accepted 18 June 2013

Published ahead of print 24 June 2013

Address correspondence to Arianna Tavanti, atavanti@biologia.unipi.it.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/AAC.00904-13.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AAC.00904-13

4314 aac.asm.org Antimicrobial Agents and Chemotherapy p. 4314–4321 September 2013 Volume 57 Number 9

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

tides (12, 13). The enhanced fungicidal activity of Hep-20observed under acidic conditions suggested that this peptidecould be used as a therapeutic agent in the topical treatment of C.glabrata vaginal infection.

However, a direct in vitro application of promising AMPs asnovel therapeutic agents has been held back by several limitations,including a significant reduction of their antimicrobial activity inthe presence of biological fluids due to physiological concentra-tions of salts, such as K�, Na�, Mg2�, and Ca2� (14–17). In addi-tion, clinical trials performed so far on AMPs have been restrictedto topical applications due to their poor bioavailability associatedwith host protease susceptibility and unknown toxicity profiles(18).

Therefore, to evaluate the potential application of Hep-20 inthe topical treatment of C. glabrata vaginitis, the peptide activitywas evaluated in (i) a vaginal fluid simulant (VFS), resemblinghuman secretion, and (ii) in human vaginal fluid (HVF), collectedfrom three healthy donors. In addition, despite the peptide previ-ously being shown to produce no significant effect on humanerythrocytes (11) and on erythroleukemia K562 human cells (19),we further investigated Hep-20 toxicity versus human peripheralblood mononuclear cells (PBMCs) and an epithelial cell line(A459) by 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetra-zolium-5-carboxanilide tetrazolium salt (XTT) reduction assayand propidium iodide (PI) staining. Indirect evidence of the pep-tide stability was also obtained by coincubating Hep-20 in thepresence of HVF.

MATERIALS AND METHODSYeast strain and growth media. A Candida glabrata clinical isolate resis-tant to fluconazole (BPY44; MIC, 256 �g/ml) was used in this study (11).BPY44 was grown in Sabouraud broth (Liofilchem s.r.l., Teramo, Italy) at30°C for 18 h. The strain was maintained on Sabouraud agar (Liofilchems.r.l.) for the duration of the study.

Antimicrobial peptide. Synthetic Hep-20 was purchased from Pep-tide Specialty Laboratories GmBH (Heidelberg, Germany). The physicalproperties of Hep-20 are listed in Table S1 in the supplemental material.Hep-20 was diluted in Milli-Q water to obtain a stock solution of 8 mg/ml.

Fungicidal activity of Hep-20 in VFS. A vaginal fluid simulant mim-icking human vaginal secretions was used to evaluate whether the activityof Hep-20 was maintained under such conditions. The VFS was preparedas described by Owen and Katz and the solution pH adjusted to 4.5 usingHCl, with the following composition: NaCl, 3.5 g/liter; KOH, 1.4 g/liter;Ca(OH)2, 0.22 g/liter; bovine serum albumin, 0.018 g/liter; lactic acid, 2g/liter; acetic acid, 1 g/liter; glycerol, 0.16 g/liter; urea, 0.4 g/liter; andglucose, 5 g/liter (20).

The solution was then filtered through a 0.22-�m filter (Merck Milli-pore, Rome, Italy) and stored at �20°C prior to use. The fungicidal activ-ity of Hep-20 against BPY44 was evaluated in a microdilution assay per-formed in VFS 4-fold-diluted in sterile deionized water to mimic averagefluid dilution induced by vaginal suppository (21).

Briefly, the C. glabrata BPY44 fluconazole-resistant isolate was grownin Sabouraud medium to exponential phase at 30°C and suspended indeionized water at a concentration of 1 � 107cells/ml. Ten microliters offungal suspension was incubated in the presence of Hep-20 (concentra-tions ranged from 25 to 100 �M) in 100 �l of VFS with and withoutEDTA, a chelator of divalent cations (1 mM final concentration).

Following a 90-min incubation, samples were diluted 10-fold in water,and 200 �l was plated onto Sabouraud plates and incubated for 24 h todetermine the CFU number. The fungicidal effect was defined as a reduc-tion in the number of viable yeast cells of �3 log10 CFU/ml compared tothat of the untreated control (11, 13).

HVF extraction. HVF was collected from three healthy donors (D1 toD3), who ranged in age from 26 to 42 years and had regular menstrualcycles of approximately 28 days. To account for protein content variabilityover the menstrual cycle, HVF was collected twice from two of the donors,before and after ovulation. The human vaginal fluid was collected as de-scribed by Valore and coworkers (22), with minor modifications: tam-pons (O.B. Pro Comfort Mini) were inserted in the vagina for 8 to10 h andthen transferred into centrifuge tubes. Thirty ml of 10 mM SPB, pH 5.0,was added to each tampon and incubated under agitation at 37°C. Thefluid retained in tampons following a 4-h incubation was squeezed out bycentrifuging at 3,800 � g to remove epithelial cells and tampon debris. Inorder to reverse the extreme dilution due to the extraction procedure, thefluid was centrifuged through a 3-kDa filter spin unit (Merck, Millipore)for 3 to 4 h up to a final volume of 1 ml. The effective protein concentra-tion of HVF was then measured by Lowry assay versus a bovine serumalbumin (BSA) standard solution (100 mg/liter BSA stock solution) (23).HVF was then diluted appropriately (final protein concentration, 2 g/li-ter) (24) and stored at �20°C until further use. The study protocol wasapproved by the local ethics committee, and all volunteers gave their in-formed written consent.

Fungicidal activity of Hep-20 in HVF. The fungicidal activity ofHep-20 against BPY44 was evaluated by killing assay performed in HVF,quantified as described above, and diluted to a protein content of 2 g/liter(24). C. glabrata BPY44 was grown in Sabouraud medium to exponentialphase at 30°C and suspended in deionized water at a concentration of 1 �107cells/ml. Fungal suspensions were incubated in the presence of Hep-20(ranging from 3.12 to 50 �M) in a 4-fold-diluted HVF in sterile deionizedwater with and without 1.5 mM EDTA (final volume of 100 �l). Followingincubation for 15 min to 24 h at 37°C, each mixture was diluted 10-fold inwater, and 200 �l was plated onto Sabouraud plates and incubated for 24h to determine CFU numbers. The fungicidal effect was defined as a re-duction in the number of viable yeast cells of �3 log10 CFU/ml comparedto the untreated control (11).

Evaluation of Hep-20 degradation in HVF. Hep-20 (5 �g) was incu-bated in 4-fold-diluted HVF from D1 and D2 for 90 min and for 24 h,respectively, at 37°C under agitation in the presence or absence of EDTA.Four-fold-diluted HVF with or without 1.5 mM EDTA was included inthe analysis as a control. Hep-20 (5 �g) incubated alone was used as anegative control of degradation. Following incubation, samples and con-trols were boiled for 10 min in SDS-loading buffer (50 mM Tris-Cl, pH6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, and 10%glycerol) and separated on 17.5% Tricine-SDS-PAGE gels (25). Molecularmass markers (7 to 175 kDa) and samples were processed for 2 h in astacking gel (4% polyacrylamide; Sigma-Aldrich) at 10 mA and for 4 h at30 mA in a separating gel. Following electrophoresis, gels were stainedwith EZ blue reagent (Sigma-Aldrich). Bands were visualized with theImage Master apparatus (VDS; GE Health Care, Milan, Italy). To evaluatethe contribution of HVF proteases to Hep-20 degradation, experimentswere performed in parallel in vaginal fluid pretreated to 100°C for 5 min toinactivate proteases (26).

PBMC isolation and A549 culturing. Heparinized venous blood ob-tained from 6 healthy volunteers was diluted in phosphate-buffered saline(PBS; Euroclone, Milan, Italy) containing 10% (vol/vol) sodium citrate(Sigma-Aldrich) and layered on a standard density gradient (Lympho-lyte-H; Cedarlane, Euroclone). Following centrifugation at 160 � g for 20min at room temperature, supernatant was removed, without disturbingthe lymphocyte/monocyte layer at the interface, in order to eliminateplatelets. The gradient was further centrifuged at 800 � g for 20 min, andperipheral blood mononuclear cells (PBMC) were collected from the in-terface. Cells were washed three times with PBS containing 0.1% (wt/vol)BSA (Sigma-Aldrich) and 10% sodium citrate (Sigma-Aldrich) and sus-pended in RPMI 1640 (Euroclone) supplemented with 2 mM L-glutamineand 10% heat-inactivated fetal calf serum (FCS; Euroclone) at a final cellconcentration of 1 � 106 cells/ml.

Human non-small-cell lung adenocarcinoma A549 cells (ATCC CCL-

Activity of Hepcidin 20 on Candida glabrata

September 2013 Volume 57 Number 9 aac.asm.org 4315

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

185; LGC Standards, Sesto San Giovanni, Italy) were cultured in tissueculture flasks in Dulbecco’s modified essential medium (DMEM) (Euro-clone) supplemented with 2 mM L-glutamine and 10% heat-inactivatedFCS. Cell cultures were maintained in a 5% CO2 atmosphere at 37°C.When cells grew to confluence, they were treated with trypsin (Sigma-Aldrich) and split in new flasks.

PI staining of PBMCs and A549 cell line incubated with Hep-20. Cellviability was analyzed by flow cytometry by evaluating PI incorporation inPBMCs and A549 cells after a 24-h exposure to Hep-20. Isolated PBMCswere seeded in a 96-well microtiter plate (1 � 105 cells/well), and Hep-20was added to wells at a final concentration of 6.25 to 100 �M. Cells incu-bated with complete medium alone served as a negative control, whilePBMCs incubated with cycloheximide (Sigma-Aldrich) (2 mg/ml finalconcentration) were used as a positive control. Plates were incubated in a5% CO2 atmosphere at 37°C for 24 h. Cells were then transferred into 12-by 75-mm polystyrene tubes and washed in Dulbecco’s PBS (D-PBS; Eu-roclone), and 5 �l of a 50 mg/liter solution of PI (Sigma-Aldrich) wasadded to each tube. Following an incubation at room temperature for 4min in the dark, 15,000 events were acquired ungated in a flow cytometer(FACSort; BD Biosciences, Mountain View, CA). CellQuest software (BDBiosciences) was used for computer-assisted analysis. In order to evaluatePI incorporation by A549, cells were detached with trypsin from a cultureflask, counted, and suspended in DMEM supplemented with 10% FCSand 2 mM L-glutamine (5 � 104 cells/ml, final concentration). Two hun-dred microliters from the suspension was seeded in a 96-well microtiterplate (1 � 104 cells/well), and cells were cultured in a 5% CO2 atmosphereat 37°C for 24 h to allow adhesion. Fifty microliters of complete mediumcontaining Hep-20 or cycloheximide at desired concentrations were thenadded to each well. Following a further 24 h of incubation, cells wereharvested by trypsinization, washed, and exposed to PI and analyzed asdescribed for PBMCs. Experiments were repeated three times. The per-centage of cytotoxicity was calculated by subtracting spontaneous PBMC/A549 cell death, measured in cultures in the absence of Hep-20 or cyclo-heximide, according to the following formula: % cytotoxicity � [(PI-positive cellssample � PI-positive cellsspontaneous)/100 � (PI-positivecellsspontaneous)] � 100.

XTT reduction assay on PBMC and A549 cell line incubated withHep-20. The effect of Hep-20 on cell viability was also assessed using acolorimetric assay based on the reduction of XTT into a water-solubleorange formazan product by mitochondrial dehydrogenases of viablecells. The XTT-PMS solution was prepared according to Scudiero andcolleagues (27). In order to improve cellular reduction of XTT, N-meth-ylphenazonium methyl sulfate (PMS), an electron coupling agent, wasused. Briefly, XTT (Sigma-Aldrich) was prepared at 1 mg/ml in pre-warmed (37°C) D-PBS and PMS at 5 mM in D-PBS (1.53 mg/ml) (Sigma-Aldrich). A 0.025 mM PMS-XTT solution was prepared by mixing 5 ml offresh XTT (1 mg/ml) and 25 �l of 5 mM PMS immediately before use andsterilized by filtration (0.22 �m; Millipore). PBMCs and A549 cells weresuspended in complete medium and seeded in a 96-well microtiter plate asdescribed above. Cells were then incubated with Hep-20 at desired con-centrations (6.25 to 100 �M final concentrations), complete mediumalone as a negative control, and Triton X-100 (Sigma-Aldrich, Milan,Italy) (0.5% [vol/vol] final concentration) as a positive control. After a24-h incubation in the presence of 5% CO2, 50 �l of the XTT-PMS solu-tion was added to each well. Following a further incubation (3 and 1 h forPBMCs and A549, respectively), the optical density at 450 nm (OD450)was measured using a microplate reader (model 550; Bio-Rad, Milan,Italy) in order to quantify the formazan production, which gave an indi-rect measure of cell viability. Experiments were performed in duplicateand repeated three times.

Statistical analysis. Statistical analysis was performed using Instatsoftware (GraphPad). One-way analysis of variance (ANOVA) was usedto evaluate the activity of Hep-20 in VFS and HVF and for cytotoxicitystudies. A P value of �0.05 was considered statistically significant.

RESULTSFungicidal activity of Hep-20 in VFS. In order to evaluatewhether the activity of Hep-20 was maintained in a biologic fluidsimulant, killing assays on C. glabrata BPY44 were performed in a4-fold-diluted artificial fluid (VFS) with Hep-20 concentrationsranging from 25 to 100 �M. Hep-20 did not produce any fungi-cidal effect (at least 3 log reduction in CFU number) at any of theconcentrations tested (data not shown), but a significant reduc-tion in BPY44 CFU number was observed only with the highestpeptide concentration (Fig. 1). To evaluate whether the reducedactivity of the peptide was due to the presence of inhibitory cat-ions, experiments were repeated in 4-fold-diluted VFS supple-mented with the chelating agent EDTA. Preliminary experimentsperformed using different EDTA concentrations (0.75 to 4 mM)indicated that 1 mM EDTA potentiated the peptide activity. Asshown in Fig. 1, the addition of 1 mM EDTA to VFS significantlyenhanced the antifungal activity of Hep-20 against C. glabrata at aconcentration of 100 �M following a 90-min incubation period(P � 0.001 by one-way ANOVA), demonstrating its fungicidalpotential.

Antifungal activity of Hep-20 in HVF. Killing experimentswere performed in 4-fold-diluted human vaginal fluid (HVF) col-lected from 3 healthy donors (D1 to D3). The total protein contentin HVF was quantified by Lowry assay, and the fluid was diluted toobtain a protein concentration of 2 g/liter. This value was chosento represent an average protein content of HVF, which was re-ported to vary between 0.018 and 3.75 g/liter (24).

When the peptide was coincubated for 90 min at 37°C with C.glabrata in the presence of HVF from D1, no antifungal activity ofHep-20 was observed up to a concentration of 50 �M. Indeed, asshown in Fig. 2A, no significant difference was observed betweenuntreated control and treated samples for all of the peptide con-centrations tested (12.5 to 50 �M). Notably, this results was notconfirmed when fluid from donor 2 was used in the killing assay.In this case, a highly significant reduction in C. glabrata CFUnumber could be evidenced in HVF (P � 0.001 by one-way

FIG 1 Killing assay of C. glabrata BPY44 in 4-fold-diluted VFS, with andwithout 1 mM EDTA, in the presence of 100 �M Hep-20. Fungal cells incu-bated in the absence of Hep-20 serve as a viable control. Data are expressedas means � standard deviations from three independent experiments.***, P � 0.001.

Del Gaudio et al.

4316 aac.asm.org Antimicrobial Agents and Chemotherapy

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

ANOVA) (Fig. 2B). Fluid from donor 3 gave an intermediate re-sult with a significant decrease in HVF fungal burden (P � 0.0003by one-way ANOVA) (Fig. 2C) following 90 min of incubation at37°C. In an attempt to restore the peptide activity, which couldhave been hampered by ion concentration, experiments were per-formed in parallel using human vaginal fluid supplemented withEDTA.

Preliminary experiments performed on 4-fold-diluted HVFand different EDTA concentrations indicated that 1.5 mM EDTApotentiated the Hep-20 activity in HVF, producing a fungicidaleffect with 50 �M Hep-20 (data not shown).

Coincubation of Hep-20 with HVF from the three donors inthe presence of 1.5 mM EDTA restored fungicidal activity follow-ing 90 min of incubation at 37°C, with a minimal fungicidal con-centration (MFC) of 25 �M (Fig. 2A and C) and 12.5 �M (Fig.2B). Although not fungicidal, even the lowest peptide concentra-tion used (12.5 �M) caused a significant reduction in BPY44 CFUnumber in vaginal fluid collected from D1 and D3 (Fig. 2A and C;P � 0.001 and P � 0.05, respectively).

Experiments were repeated using vaginal fluid from donors 1and 2, collected before and after ovulation. For each of the twodonors, the results obtained in HVF collected before ovulation did

not significantly differ from those obtained with HVF collectedpostovulation (data not shown).

Time-killing assay of 12.5 and 25 �M Hep-20 versus BPY44performed in HVF from donor 1 confirmed that the highest pep-tide concentration exerts a fungicidal effect following 90 min ofincubation in the presence of 1.5 mM EDTA (Fig. 3). Notably, astatistically significant reduction in fungal cell number (P � 0.05)was observed starting at 15 min of incubation with 25 �M Hep-20and at 30 min with 12.5 �M Hep-20 (Fig. 3). When coincubationwas prolonged to 24 h, Hep-20 produced a fungicidal effect at alower concentration as well (12.5 �M) (Fig. 3).

Evaluation of Hep-20 degradation in HVF. The stability ofHep-20 was evaluated by coincubating the peptide in the presenceof HVF collected from two healthy donors (D1 and D2). Thepeptide was incubated in the presence of 4-fold-diluted HVF for90 min and for 24 h at 37°C. In parallel, the same experiment wasperformed on HVF preheated at 100°C for 5 min to inactivateproteases in the fluid. Samples, including HVF plus Hep-20 withand without 1.5 mM EDTA, HVF plus EDTA, Hep-20 alone (5�g), and Hep-20 alone incubated for 90 min and for 24 h, wereseparated in a 17.5% Tricine SDS-PAGE gel. As shown in Fig. 4Aand B, no band corresponding to the molecular mass of Hep-20

FIG 2 Fungicidal activity of Hep-20 against C. glabrata BPY44 in 4-fold-diluted HVF from donor 1 (D1) (A), donor 2 (D2) (B), and donor 3 (D3) (C). Fungalcells incubated in the absence of Hep-20 serve as a viable control (CTRL). Data are expressed as means � standard deviations from at least three independentexperiments. The arrows indicate fungicidal effect. NS, not significant. ***, P � 0.001; **, P � 0.01; *, P � 0.05.

Activity of Hepcidin 20 on Candida glabrata

September 2013 Volume 57 Number 9 aac.asm.org 4317

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

(lanes 1 and 2) could be observed in 4-fold-diluted HVF (Fig. 4Aand B, lanes 7 and 8). Coincubation for 90 min and for 24 h at 37°Cof the peptide with HVF from D1 (Fig. 4A and B, lanes 3 and 4)resulted in a partial degradation of Hep-20 compared to that of thepeptide alone (Fig. 4A and B, lanes 1 and 2). Preheating of HVFresulted in a complete lack of peptide degradation (Fig. 4A and B,lanes 5 and 6). The same results were obtained with HVF fromdonor 2, with a similar extent of peptide degradation (data notshown).

Propidium iodide staining. The potential cytotoxic activity ofHep-20 was assessed by PI staining of PBMC and A549. To thisaim, PBMC and A549 cell suspensions were incubated for 24 h inthe presence of Hep-20 at concentrations ranging from 6.25 to 100�M and in the presence of 2 mg/ml cycloheximide as a positive

(death) control. PBMC and A549 cellular suspensions were coin-cubated with PI for 4 min and then processed by flow cytometry.Hep-20 produced a negligible cytotoxic effect at all of the concen-trations tested on PBMC and A549, with more than 90% viablecells posttreatment.

XTT reduction assay. The cytotoxicity of Hep-20 on PBMCand on a human-derived epithelial cell line (A549) was also as-sessed by XTT assay. PBMC and A549 cell suspensions were incu-bated for 24 h in the presence of different concentrations ofHep-20 (6.25 to 100 �M) and in the presence of 0.5% Triton-X asa positive (death) control. Following this incubation, PBMC andA549 cellular suspensions were coincubated with XTT for 3 and 1h, respectively, at 37°C, and OD450 values provided informationon cell metabolic activity. Notably, none of the Hep-20 concen-trations tested (6.25 to 100 �M) showed any cytotoxic effect onPBMC and A549 cells (Fig. 5A and B). Instead, a stimulatory effectcould be observed at the highest peptide concentration tested,which was statistically significant in the case of A549 (Fig. 5A; P �0.0033 by one-way ANOVA).

DISCUSSION

Candida glabrata vaginitis is difficult to treat due to an intrinsicreduced susceptibility to azoles, in particular to fluconazole,which is widely used for the topical treatment of this mycosis (1,28). For this reason, there is an urgent need to individuate novelstrategies for the treatment of these mycoses. In this regard, thetherapeutic potential of natural anti-infective agents, such asAMPs, has generated much interest due to the potential therapeu-tic approaches that these molecules offer (5).

The present study was focused on human Hep-20, which wasdemonstrated to exert antimicrobial activity against clinically rel-evant Gram-positive and Gram-negative bacterial species (13).

We have recently shown that Hep-20 produced a fungicidaleffect on C. glabrata clinical isolates with different levels of flu-conazole susceptibility (11). This finding was particularly interest-ing in view of the scarce susceptibility this species presents toAMPs as well as to azoles (3, 7–10). Furthermore, the enhancedfungicidal activity of Hep-20 observed under acidic conditionssuggested that this peptide could be used as a therapeutic agent ofC. glabrata-related infections in body regions characterized by lowpH, such as the vaginal region (11). No data are currently availablewith regard to the mode of action of Hep-20 on C. glabrata. Indi-rect information can be extrapolated from a recent study based on

FIG 3 Time-kill curves of Hep-20 (12.5 and 25 �M) against C. glabrata BPY44 in 4-fold-diluted HVF from D1 with and without 1.5 mM EDTA. Fungal cellsincubated in the absence of Hep-20 served as a viable control (CTRL). Data are expressed as means � standard deviations from three independent experiments.The arrows indicate fungicidal effect. *, P � 0.05; **, P � 0.01.

FIG 4 SDS-PAGE gels of the peptide alone and coincubated in the presence ofHVF from donor D1 following 90 min in HVF at 37°C (A) and 24 h of incu-bation in HVF at 37°C (B). Lane M, molecular marker (7–175 kDa); 1, Hep-20(5 �g); 2, Hep-20 (5 �g) plus 1.5 mM EDTA; 3, HVF plus Hep-20; 4, HVF plus1.5 mM EDTA plus Hep-20; 5, HVF (preheated at 100°C for 5 min) plusHep-20; 6, HVF (preheated at 100°C for 5 min) plus Hep-20 plus 1.5 mMEDTA; 7, HVF plus 1.5 mM EDTA; 8, HVF (preheated at 100°C for 5 min) plus1.5 mM EDTA.

Del Gaudio et al.

4318 aac.asm.org Antimicrobial Agents and Chemotherapy

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

the interaction between Hep-20 and Escherichia coli (29). Thisstudy demonstrates a greater tendency of the peptide to destabilizethe E. coli membrane under acidic conditions than at neutral pHthrough formation of pores, as evidenced by the release of �-ga-lactosidase from Hep-20 treated cells at pH 5.0 and formation ofblebs (29).

To verify the applicability of the peptide in the vaginal environ-ment, the antifungal activity of Hep-20 was measured in an arti-ficial vaginal fluid simulating human vaginal secretions (VFS).Considering that the average volume of vaginal fluid ranges from0.5 to 0.75 ml (21), in our experiments VFS was diluted 4-fold tomimic fluid dilution by a vaginal suppository (20, 21).

In this experimental condition, the Hep-20 fungicidal effectpreviously observed in SPB (11) was not confirmed for any of theconcentrations tested (25 to 100 �M). This finding suggested thatHep-20 activity could be inhibited by the presence in the mediumof VFS components, such as BSA, and mono- and divalent cations,such as Ca2�, K�, and Na� (20). Indeed, the antimicrobial activityof several AMPs has been shown to be markedly reduced in bio-logical fluids due to the presence of proteins, other components,or physiological concentration of salts, such as K�, Na�, Mg2�,and Ca2� (14–17).

Among the different mechanisms proposed to explain the lossof antimicrobial activity of AMPs in biological fluids, it has beenhypothesized that the presence of divalent cations, such as calciumand magnesium, mediates an electrostatic interaction with thenegatively charged membrane of microorganisms. This creates abarrier preventing the interaction between AMPs and microbialplasma membrane. In an attempt to restore the activity of thepeptide in VFS, experiments were repeated in the presence of 1

mM EDTA. This synthetic chelating agent forms strong com-plexes with cations, and it is widely used as a stabilizer in food andin pharmaceutical vaginal applications (30, 31). In vitro studieshave demonstrated that prolonged exposure of ectocervical tissueto EDTA can produce a toxic effect at concentrations higher than2.5 mM (31). Furthermore, EDTA has been demonstrated to in-hibit hyphal development in C. albicans and to exert antifungalactivity against Saccharomyces cerevisiae at concentrations of 10and 17 mM, respectively (30, 32). In a study performed on humansaliva by Wei and Bobek, the addition of EDTA enhanced theantifungal activity of different peptides, such as MUC7, histatin 5,and magainin, against C. albicans compared to saliva alone (14).

In our experiments, the addition of EDTA to the VFS resultedin a significant enhancement of Hep-20 fungicidal activity againstC. glabrata BPY44 following 90 min of incubation at 30°C. Thisfinding reinforced the view that divalent cations inhibit the activ-ity of Hep-20: the presence of EDTA most likely enhanced Hep-20activity by removing cations that protect the cell membrane frominteraction with the peptide or by sequestering ions which arerequired by fungal intracellular enzymes (14, 30).

This finding was particularly interesting in view of a potentialapplication of this peptide in the topical treatment of fungal vag-initis. To further pursue this possibility, the fungicidal activity ofHep-20 was evaluated in HVF collected from 3 healthy donors(D1 to D3) and diluted to a final protein concentration of 2 g/literto consider HVF biological variability (24).

Killing experiments in HVF indicated that a biological variabil-ity occurred among vaginal fluid collected from three differentdonors, since for two of them a significant reduction in the num-ber of viable fungal cells was induced by Hep-20 in the fluid alonefollowing 90 min of incubation at 37°C. However, in the case ofD1, no antifungal activity was observed in HVF in the absence ofEDTA. This finding indicates that vaginal fluid composition dif-ferentially affects the peptide activity. Indeed, it is recognized thatHVF contains several AMPs, such as calcoprotectin, lysozyme,SLPI, HBD1-2, HNP1-3, and LL-37, which could act synergisti-cally with Hep-20, overcoming the natural inhibitory effectcaused by ions or other fluids components (22, 26, 33). Notably,when 1.5 mM EDTA was added to the HVF, Hep-20 exerted acomplete fungicidal effect versus C. glabrata BPY44 following 90min of incubation, and it was maintained after 24 h of incubation.Interestingly, the Hep-20 MFC that produced a fungicidal effectwas 25 �M for all donors, a quarter of the peptide concentrationthat significantly reduced fungal cell number in VFS. In addition,experiments performed using HVF collected before and after ovu-lation gave similar results for each of the donors considered, sug-gesting that variation of HVF composition during the menstrualcycle does not significantly influence Hep-20 activity. Overall, theresults obtained on Hep-20 fungicidal activity in VFS and, moreimportantly, in HVF reinforced the view that this peptide is apromising candidate for topical treatments of C. glabrata vaginalinfections.

In order to assess the potential of Hep-20 as a therapeutic agentagainst C. glabrata vaginal infections, we investigated whether thepeptide retains its activity in physiological conditions. The resultsobtained in this study demonstrated that Hep-20 maintains itsantifungal activity in the presence of EDTA at pH, salt concentra-tion, and protein content similar to those found in vivo (VFS).Moreover, this finding was also confirmed when the peptide ac-tivity was evaluated in HVF. At present, no data are available with

FIG 5 OD values (at 450 nm) of formazan salt production by A549 cells (A)and PBMCs (B) incubated with concentrations of Hep-20 ranging from 6.25 to100 �M. Triton X-100 (0.5%, vol/vol) served as a positive control. The vitalitypercentage, shown for each condition, was calculated by considering the neg-ative control as 100% viability. Data are expressed as means from three sepa-rate experiments � standard errors of means. **, P � 0.0033.

Activity of Hepcidin 20 on Candida glabrata

September 2013 Volume 57 Number 9 aac.asm.org 4319

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

regard to the stability of the peptide in HVF. Indirect evidence ofHep-20 stability was obtained in this study from coincubationexperiments of the peptide with HVF at 37°C at two different timepoints, 90 min and 24 h. SDS-PAGE gels indicated that when thepeptide was coincubated with HVF left unsupplemented or sup-plemented or with 1.5 mM EDTA, Hep-20 was partially degradedat the two time points considered. However, when the peptide wascoincubated with preheated HVF, Hep-20 gave a band similar inintensity to that of the untreated peptide, suggesting that the par-tial degradation observed was mainly due to fluid trypsin-like pro-teases. Indeed, a proteomic analysis on cervicovaginal fluids iden-tified several proteases, such as serine (e.g., kallikreins) andcysteine (e.g., cathepsins) ones, which could degrade Hep-20 (26,39). Further experiments will be required to understand the exactmechanism leading to this partial peptide degradation following90 min or 24 h of incubation in HFV. Nevertheless, killing exper-iments performed in HVF clearly demonstrated that, even if par-tially degraded, the peptide is still able to exert fungicidal activityagainst C. glabrata at both 90 min and within 24 h. However, thepartial degradation of the peptide may be overcome by synthesiz-ing hepcidin-derived peptides containing D-amino acids and/orby incorporating these peptide analogues in nanoparticles, en-hancing the bioavailability of Hep-20 in HVF.

Even if Hep-20 is a human-derived peptide, therapeuticallyactive concentrations can be higher than physiological ones.Therefore, it is mandatory to investigate the potential cytotoxiceffects exerted by this peptide on human cells. Previous data re-port a lack of hemolytic effect on human erythrocytes (11).

In this study, we expanded upon the Hep-20 cytotoxicity ex-periments by evaluating the effect of different peptide concentra-tions on human PBMC and a human-derived epithelial cell line(A549) by PI staining and XTT reduction assay. The results ob-tained showed that all Hep-20 concentrations tested (6.25 to 100�M) produced negligible or no cytotoxic effect on PBMC andA549 cells, as demonstrated by PI staining and XTT assay. Theseresults are in agreement with a study by Park and coworkers inwhich Hep-20 was tested for cytotoxicity on erythroleukemia typeK562 human cells at concentrations of up to 30 �M (about 3,000-fold higher than that found in urine), with no significant effect onhuman cells (88% viable cells post treatment) (19).

It is worth underlining that Hep-20 produced a significant in-crease in A549 cell metabolism at the highest concentration tested(100 �M). This result is not surprising, since it is reported thatsome antimicrobial peptides, such as hBD-2, are able to induceA549 proliferation in a dose-dependent manner (34). In addition,the role of AMPs in wound healing is supported by data on theactivity of cathelicidin LL-37, hBD-2, and hBD-3 (35–37). Thesepeptides are highly expressed in epidermal keratinocytes in re-sponse to injury or infection of the skin (35–37). It has also beenshown that treatment with exogenous hBD-3 leads to an enhancedre-epithelialization of wounds in animal models (37). Hep-20shares a high structural homology with defensins, and it couldexert a similar stimulatory effect on A549 cells (38).

In conclusion, despite the fact that further studies are stillneeded to characterize the peptide mode of action and to optimizeHep-20 fungicidal activity, our results indicate that this peptide orits derivatives should be further studied as novel therapeuticagents for the control of vaginal C. glabrata infections.

ACKNOWLEDGMENT

Financial support was obtained from Fondi di Ateneo, University of Pisa.

REFERENCES1. Kennedy MA, Sobel JD. 2010. Vulvovaginal candidiasis caused by non-

albicans Candida species: new insights. Curr. Infect. Dis. Rep. 12:465– 470.2. Richter SS, Galask RP, Messer SA, Hollis RJ, Diekema DJ, Pfaller MA.

2005. Antifungal susceptibilities of Candida species causing vulvovaginitisand epidemiology of recurrent cases. J. Clin. Microbiol. 43:2155–2162.

3. Pfaller MA. 2012. Antifungal drug resistance: mechanisms, epidemiology,and consequences for treatment. Am. J. Med. 125(Suppl. 1):S3–S13.

4. Ostrosky-Zeichner L. 2013. Candida glabrata and FKS mutations: wit-nessing the emergence of the true Multidrug-Resistant Candida. Clin.Infect. Dis. 56:1733–1734.

5. Hancock RE, Sahl H-G. 2006. Antimicrobial and host-defense peptides asnew anti-infective therapeutic strategies. Nat. Biotechnol. 24:1551–1557.

6. Zasloff M. 2002. Antimicrobial peptides of multicellular organisms. Na-ture 415:389 –395.

7. Joly S, Maze C, McCray PB, Jr, Guthmiller JM. 2004. Human beta-defensins 2 and 3 demonstrate strain-selective activity against oral micro-organisms. J. Clin. Microbiol. 42:1024 –1029.

8. Helmerhorst EJ, Venuleo C, Beri A, Oppenheim FG. 2005. Candidaglabrata is unusual with respect to its resistance to cationic antifungalproteins. Yeast 22:705–714.

9. Helmerhorst EJ, Venuleo C, Sanglard D, Oppenheim FG. 2006. Roles ofcellular respiration, CgCDR1, and CgCDR2 in Candida glabrata resistanceto histatin 5. Antimicrob. Agents Chemother. 50:1100 –1103.

10. Benincasa M, Scocchi M, Pacor S, Tossi A, Nobili D, Basaglia G, BusettiM, Gennaro R. 2006. Fungicidal activity of five cathelicidin peptidesagainst clinically isolated yeasts. J. Antimicrob. Chemother. 58:950 –959.

11. Tavanti A, Maisetta G, Del Gaudio G, Petruzzelli R, Sanguinetti M,Batoni G, Senesi S. 2011. Fungicidal activity of the human peptide hep-cidin 20 alone or in combination with other antifungals against Candidaglabrata isolates. Peptides 32:2484 –2487.

12. Kacprzyk L, Rydengård V, Mörgelin M, Davoudi M, Pasupuleti M,Malmsten M, Schmidtchen A. 2007. Antimicrobial activity of histidine-rich peptides is dependent on acidic conditions. Biochim. Biophys. Acta1768:2667–2680.

13. Maisetta G, Petruzzelli R, Brancatisano FL, Esin S, Vitali A, Campa M,Batoni G. 2010. Antimicrobial activity of human hepcidin 20 and 25against clinically relevant bacterial strains: effect of copper and acidic pH.Peptides 31:1995–2002.

14. Wei GX, Bobek LA. 2005. Human salivary mucin MUC7 12-mer-L and12-mer-D peptides: antifungal activity in saliva, enhancement of activitywith protease inhibitor cocktail or EDTA, and cytotoxicity to human cells.Antimicrob. Agents Chemother. 49:2336 –2342.

15. Batoni G, Maisetta G, Esin S, Campa M. 2006. Human beta-defensin-3:a promising antimicrobial peptide. Mini Rev. Med. Chem. 6:1063–1073.

16. Wei GX, Campagna AN, Bobek LA. 2007. Factors affecting antimicrobialactivity of MUC7 12-mer, a human salivary mucin-derived peptide. Ann.Clin. Microbiol. Antimicrob. 6:14.

17. Maisetta G, Di Luca M, Esin S, Florio W, Brancatisano FL, Bottai D,Campa M, Batoni G. 2008. Evaluation of the inhibitory effects of humanserum components on bactericidal activity of human beta defensin 3.Peptides 29:1– 6.

18. Yeung AT, Gellatly SL, Hancock RE. 2011. Multifunctional cationic hostdefence peptides and their clinical applications. Cell. Mol. Life Sci. 68:2161–2176.

19. Park CH, Valore EV, Waring AJ, Ganz T. 2001. Hepcidin, a urinaryantimicrobial peptide synthesized in the liver. J. Biol. Chem. 276:7806 –7810.

20. Owen DH, Katz DF. 1999. A vaginal fluid stimulant. Contraception59:91–95.

21. Lai BE, Xie YQ, Lavine ML, Szeri AJ, Owen DH, Katz DF. 2008.Dilution of microbicide gels with vaginal fluid and semen simulants: effecton rheological properties and coating flow. J. Pharm. Sci. 97:1030 –1038.

22. Valore EV, Park CH, Igreti SL, Ganz T. 2002. Antimicrobial componentsof vaginal fluid. Am. J. Obstet. Gynecol. 187:561–568.

23. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. 1951. Protein mea-surement with the Folin phenol reagent. J. Biol. Chem. 193:265–275.

24. Tomás MS, Nader-Macías ME. 2007. Effect of a medium simulatingvaginal fluid on the growth and expression of beneficial characteristics of

Del Gaudio et al.

4320 aac.asm.org Antimicrobial Agents and Chemotherapy

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

potentially probiotic lactobacilli, p 732–739. In Méndez-Vilas A (ed),Communicating current research, educational topics and trends in ap-plied microbiology. Formatex, Badajoz, Spain.

25. Schägger H, von Jagow G. 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in therange from 1 to 100 kDa. Anal. Biochem. 166:368 –379.

26. Shaw JL, Smith CR, Diamandis EP. 2007. Proteomic analysis of humancervico-vaginal fluid. J. Proteome Res. 6:2859 –2865.

27. Scudiero DA, Shoemaker RH, Paull KD, Monks A, Tierney S, NofzigerTH, Currens MJ, Seniff D, Boyd MR. 1988. Evaluation of a solubletetrazolium/formazan assay for cell growth and drug sensitivity in cultureusing human and other tumor cell lines. Cancer Res. 48:4827– 4833.

28. Sanguinetti M, Posteraro B, Fiori B, Ranno S, Torelli R, Fadda G. 2005.Mechanism of azole resistance in clinical isolates of Candida glabrata col-lected during a hospital survey of antifungal resistance. Antimicrob.Agents Chemother. 49:668 – 679.

29. Maisetta G, Vitali A, Scorciapino MA, Rinaldi AC, Petruzzelli R, Bran-catisano FL, Esin S, Stringaro A, Colone M, Luzi C, Bozzi A, Campa M,Batoni G. 2013. pH-dependent disruption of Escherichia coli ATCC25922 and model membranes by the human antimicrobial peptides hep-cidin 20 and 25. FEBS J. 280:2842–2854.

30. Kubo I, Lee SH, Ha TJ. 2005. Effect of EDTA alone and in combinationwith polygodial on the growth of Saccharomyces cerevisiae. J. Agric. FoodChem. 53:1818 –1822.

31. Sassi AB, Cost MR, Cole AL, Cole AM, Patton DL, Gupta P, Rohan LC.2011. Formulation development of retrocyclin 1 analog RC-101 as ananti-HIV vaginal microbicide product. Antimicrob. Agents Chemother.55:2282–2289.

32. Gil ML, Casanova M, Martínez JP. 1994. Changes in the cell wall glyco-protein composition of Candida albicans associated to the inhibition ofgerm tube formation by EDTA. Arch. Microbiol. 161:489 – 494.

33. Singh PK, Tack BF, McCray PB, Jr, Welsh MJ. 2000. Synergistic andadditive killing by antimicrobial factors found in human airway surfaceliquid. Am. J. Physiol. Lung Cell. Mol. Physiol. 279:L799 – 805.

34. Zhuravel E, Shestakova T, Efanova O, Yusefovich Y, Lytvin D, Soldat-kina M, Pogrebnoy P. 2011. Human beta-defensin-2 controls cell cycle inmalignant epithelial cells: in vitro study. Exp. Oncol. 33:114 –120.

35. Dorschner RA, Pestonjamasp VK, Tamakuwala S, Ohtake T, Rudisill J,Nizet V, Agerberth B, Gudmundsson GH, Gallo RL. 2001. Cutaneousinjury induces the release of cathelicidin anti-microbial peptides activeagainst group A Streptococcus. J. Investig. Dermatol. 117:91–97.

36. Sørensen OE, Cowland JB, Theilgaard-Mönch K, Liu L, Ganz T, Bor-regaard N. 2003. Wound healing and expression of antimicrobial pep-tides/polypeptides in human keratinocytes, a consequence of commongrowth factors. J. Immunol. 170:5583–5589.

37. Hirsch T, Spielmann M, Zuhaili B, Fossum M, Metzig M, Koehler T,Steinau HU, Yao F, Onderdonk AB, Steinstraesser L, Eriksson E. 2009.Human beta-defensin-3 promotes wound healing in infected diabeticwounds. J. Gene Med. 11:220 –228.

38. Jordan JB, Poppe L, Haniu M, Arvedson T, Syed R, Li V, Kohno H, KimH, Schnier PD, Harvey TS, Miranda LP, Cheetham J, Sasu BJ. 2009.Hepcidin revisited, disulfide connectivity, dynamics, and structure. J. Biol.Chem. 284:24155–24167.

39. Moncla BJ, Pryke K, Rohan LC, Graebing PW. 2011. Degradation ofnaturally occurring and engineered antimicrobial peptides by proteases.Adv. Biosci. Biotechnol. 2:404 – 408.

Activity of Hepcidin 20 on Candida glabrata

September 2013 Volume 57 Number 9 aac.asm.org 4321

on March 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from